The overall aim of this project to determine, using 31-Phosphorus nuclear magnetic resonance spectorscopy, if brain metabolic markers exist for subcortic vascular dementia and a contrasting degenerative dementia. If present, would these markers distinguish ischemic dementia patients from Alzheimer disease patients and from healthy, elderly controls? A five year longitudinal and cross- sectional study will noninvasively investigate clinical, physiologic (regional cerbral blood flow), anatomic (magnetic resonance imaging), metabolic (in vivo 31PNMR) and pathologic (autopsy, post mortem in vitro 31PNMR data in giggly selected, strictly diagnosed patients with multiple subcortical ischemic dementia, Alzheimer's disease and healthy, elderly controls to characterize changes in brain energy phosphate, phospholipid metabolic studies have failed to identify specific features or markers for vascular and degenerative dementia. Further, markers donot currently exist to exclude, antemortem, Alzheimer's changers in patients with vascular dementia. Because the differential diagnosis of ischemic and degenerative dementia may be difficult, and because vascular caused are potentially preventable, capable of being arrested and reversible, the ability to develop and apply an accurate noninvasive method to help identify and differentiate these groups of patients is important.